The proposed research develops and pilot-tests a novel, theoretically and empirically derived treatment aimed at facilitating change in joint attention interactions between caregivers and their toddlers with autism. Young children with autism show impairments in engaging in joint attention with others. The impairment affects their ability to sustain a shared interest in a dyadic interaction and to use specific joint attention skills such as pointing and showing. The importance of joint attention is underscored by data suggesting these skills as important to later language. Yet these interaction and skill deficits have rarely been the focus of systematic intervention efforts, particularly with caregivers. Thus, targeting these deficits in developmentally young children using familiar caregivers may result in better child language outcomes. We will pilot-test the intervention in 3 phases. Each phase will include 10 caregiver-child dyads with 5 randomized to begin an 8 week, 3 day per week hourly intervention and the other 5 wait-listed to begin the treatment 12 weeks later for a total of 30 caregiver-child dyads. Each dyad will contain a child with autism who is under 36 months of age and under 20 months expressive language age. The intervention will focus on teaching caregivers to initiate and maintain episodes of joint engagement with their children. Caregiver and child joint attention behaviors will be assessed over the course of intervention as well as before and after treatment. Thus, the intervention is tested in a multi-assessment, three-cohort design with decision point breaks in-between each phase of intervention. Thus, the intervention can be continually assessed and modified as needed. The goals of the pilot intervention study are to develop a manualized treatment protocol, to pilot test a series of caregiver-child dyads and then assess responses of both caregiver and child to the treatment. The exploratory and pilot research will lay the necessary groundwork leading up to a full-scale randomized and comparative study of the treatment.